As generative AI takes on more creative decision-making, authorship and authority in the art world will shift — blurring credit, altering valuations, and concentrating rents among platforms and model owners, unless institutions and policy adapt.

Main Finding

The paper argues that as AI systems are increasingly integrated into artistic practice, artistic worldbuilding will shift along two dimensions — artistic autonomy (who/what makes creative decisions) and artistic authority (who/what sets norms and meaning) — producing distinct human–AI interaction scenarios. By using qualitative foresight and scenario analysis, the authors show a plausible trajectory from current generative-art practices (human-led design + AI generation) toward futures where AI assumes larger creative and institutional roles, requiring new ways to recognise and value diverse artistic contributions (from human inspiration and curation to non-human decision-making).

Key Points

• Framing: distinguishes macro art worlds (art societies, institutions, markets) from micro artworlds (the worlds created inside individual artworks) and analyses human–AI relations across both scales.
• Core conceptual axes: artistic autonomy (who executes creative decisions) and artistic authority (who defines meaning, norms, distribution of credit).
• Three scenarios (conceptual):
• Current generative-art paradigm — humans design systems, craft prompts, curate outputs; AI functions as a tool/medium with human-directed constraints and final human curation/selection.
• Hybrid co-creation — AI takes on substantive decision-making inside artworks (micro worldbuilding) while humans retain curatorial/authoritative roles at the macro level; authorship and credit become shared/blurred.
• AI-led worldbuilding — AI systems autonomously generate both artefacts and the institutional meanings or practices (macro-level), pushing humans toward roles of audience, regulator, or meta-curator.
• The scenarios serve as a theoretical test bed to classify types of contributions (inspiration, curation, automation, autonomous generation) and to anticipate sociocultural and institutional consequences.
• The paper is theoretical and exploratory; it aims to surface conceptual distinctions and normative questions rather than provide empirical validation.

Data & Methods

• Methodology: qualitative foresight / scenario-building approach.
• Evidence base: literature review of generative art practices and theoretical work on artistic authorship, autonomy, and worldbuilding.
• Analysis: conceptual mapping of roles and interactions across the two axes (autonomy and authority) and construction of three hypothetical scenarios illustrating progressive shifts in AI involvement.
• Limitations: no empirical or quantitative data; scenarios are normative/hypothetical and intended to provoke theoretical reflection and further empirical work.

Implications for AI Economics

• Labor and complementarities
  • Changing division of labor: as AI assumes more decision-making, demand for certain human creative tasks (prompt engineering, curation, conceptual leadership) may rise while routine or executional creative labor declines.
  • Complementarity opportunities: human skills (taste, contextual judgment, institution-building) may complement AI capabilities, creating new premium roles and tasks.
• Valuation and markets
  • Attribution and quality signals: blurred authorship complicates provenance, signaling, and valuation — markets may discount AI-generated works or create new quality tiers (human-authored premium vs AI-authored).
  • Product differentiation: AI-driven worldbuilding can dramatically increase supply and diversity of cultural goods, affecting pricing dynamics, long-tail markets, and consumer attention competition.
• Intellectual property and revenue models
  • IP allocation and royalties: when AI contributes non-trivial creative decisions, legal and contract frameworks must adapt (royalty sharing, new licensing forms, rights for model outputs).
  • Platforms and intermediaries: platform providers and model owners may capture rents (data, models, distribution), increasing concentration and gatekeeping in art markets.
• Market structure and investment incentives
  • Capitalization of AI as creative capital: firms and institutions investing in autonomous creative systems could realize economies of scale and network effects, reshaping competitive dynamics in creative industries.
  • Public goods and externalities: widespread use of shared AI models may generate cultural commons (style diffusion) or negative externalities (devaluation of certain artistic labor), suggesting roles for policy intervention.
• Measurement and policy
  • Measurement challenges: GDP and labor statistics may undercount creative contributions when AI is involved; new metrics needed to capture hybrid human–AI creative output and welfare effects.
  • Policy levers: regulation on attribution, transparency of model involvement, data provenance, and competition policy for platforms/models could materially shape incentives and distributional outcomes.
• Research opportunities for AI economists
  • Empirically test scenario outcomes: measure how valuation, wages, and market shares change under varying degrees of AI autonomy and authority.
  • Study complementarities vs substitution across creative tasks and the resulting wage/productivity effects.
  • Model platform rents and two-sided market dynamics in AI-mediated art ecosystems.
  • Analyze optimal IP and contract designs for mixed human–AI authorship and licensing.

Overall, the paper highlights institutional and normative dimensions of AI-driven cultural production that have direct economic implications: changing labor divisions, new forms of market concentration, evolving property rights, and measurement/policy challenges. These issues merit empirical and theoretical attention from AI economists to understand distributional effects and to design policies that preserve creative value and fair compensation.